The Risks of “AI Washing”: What Businesses Need to Know
AI – artificial intelligence – is a buzzword that has been prone to get caught up in hype and hyperbole. A company that communicates to its investors using hype and hyperbole about its AI capabilities runs the same risk as any company that embellishes the truth.
We now know that when there is nothing behind the hype – when AI actually is not being used by the company in the way that it claims – the Securities and Exchange Commission (SEC), which regulates the securities markets to protect investors, might call the company on it. Specifically, the SEC may bring an enforcement action that can entail heavy fines and bad publicity.
The SEC’s intention to police what is known as “AI Washing” is vividly demonstrated by the recent enforcement actions against two investment advisers, Delphia (USA) Inc. and Global Predictions Inc., described below in greater detail.
What is AI Washing?
AI washing refers to the practice of making inflated or false claims about a company’s AI capabilities, often to gain a competitive edge, attract investment, or mislead consumers. Like “greenwashing,” in which companies overstate their environmental credentials, AI washing can lead to a distorted market perception and erode trust in legitimate AI innovators.
However, AI washing differs from greenwashing in some key aspects. AI technologies are often more complex and less tangible than environmental practices, making it harder for stakeholders to assess the veracity of AI-related claims. Additionally, the rapid pace of AI development means that regulatory frameworks and industry standards are still evolving, creating uncertainty around what constitutes deceptive AI marketing.
AI washing and also lead to a violation of various provisions of the federal securities laws. Section 18 of the Exchange Act of 1934 imposes liability for false and misleading statements in documents filed with the SEC to any person who makes such false or misleading statements. Section 10(b) and Rule 10b-5 of the Exchange Act broadly prohibit fraudulent and deceptive practices and untrue statements or omissions of material facts in connection with the purchase or sale of any security. Section 20 of the Exchange Act and Section 15 of the Securities Act of 1933 provide that a person controlling any person liable under those statutes may be liable jointly and severally and to the same extent as its controlled person for violations of the Exchange Act or the Securities Act. Almost all state securities laws provide for similar prohibitions of and liabilities for false and misleading statements.
Recent SEC Enforcement Actions
Toronto-based Delphia (USA) Inc. faced charges for making false and misleading statements on its web site, in a press release and in SEC filings about its use of AI and machine learning in its investment process, claiming to use AI to predict successful investments. The SEC found these statements to be untrue, as Delphia did not possess the claimed AI capabilities.
Similarly, Global Predictions Inc. falsely claimed on its website and in social media to be the “first regulated AI financial advisor” and misrepresented its platform’s “expert AI-driven forecasts.” The SEC claimed these statements were not true.
The companies agreed to settle the SEC’s charges, paying civil penalties of $225,000 and $175,000 respectively without admitting culpability.
With companies fiercely competing to align themselves with AI even as AI’s possibilities for business are expanding faster than many companies can integrate these new developments into their products and services, it is unlikely that these will be the last AI washing enforcement actions by the SEC. The fate of Delphi and Global Predictions should serve as a warning to businesses about the consequences of making false AI claims.
Legal Risks Beyond SEC Enforcement
In addition to SEC enforcement, companies engaging in AI washing also face the risk of private lawsuits from investors, consumers, and other stakeholders. These claims can be brought under various federal and state laws, including securities laws, consumer protection statutes, and common law fraud.
For example, investors may bring claims under the Securities Act or the Exchange Act of 1934 for material misstatements or omissions related to a company’s AI capabilities. Consumers may also pursue claims under state unfair competition and false advertising laws, arguing that they were misled by a company’s AI washing practices.
Mitigating AI Washing Risks
To minimize the legal and reputational risks associated with AI washing, businesses should adopt the following best practices:
-
-
- Ensure accuracy and transparency in AI-related claims: All statements about a company’s AI capabilities should be factual, precise, and substantiated. If AI technologies are limited or still in development, this should be clearly disclosed to avoid misleading stakeholders.
- Implement robust internal controls: Establish policies and procedures for reviewing and approving AI-related claims, conduct regular audits of AI systems and disclosures, and train employees on responsible AI communication.
- Stay informed about regulatory developments: Monitor SEC guidance, enforcement actions, and rule-making initiatives related to AI disclosures, and stay attuned to evolving industry standards and best practices.
- Consult with legal counsel: Work closely with experienced attorneys who can provide guidance on AI-related disclosures, help navigate the evolving regulatory landscape, and advise on best practices for avoiding AI washing claims.
-
Conclusion
As the AI landscape continues to evolve, businesses must be vigilant in avoiding the pitfalls of AI washing. By prioritizing accuracy, transparency, and compliance in their AI-related claims, companies can harness the power of AI while mitigating legal and reputational risks. The SEC’s recent enforcement actions serve as a stark reminder of the consequences of engaging in AI washing and underscore the importance of seeking expert legal guidance in this rapidly developing area.
Sorry, the comment form is closed at this time.